Abstract
Background: 60 000 women in Norway give birth each year. Out of these 10-15% suffer from perinatal depression. A large body of research has shown the significant impact of mothers’ mental health on mother-child-attachment. Children’s temperamental style also contribute to the quality of the attachment. Attachment is proven to be an important factor in both mother and child development and health, with lifelong repercussions. In this paper I investigate whether a Norwegian web-based intervention (Mamma Mia) targeted at perinatal women’s mental health, increases mothers’ perceived attachment towards their babies. The main hypothesis is that women in the intervention group will have healthier attachment scores compared to controls. It is further examined whether the development of attachment over time in the two groups is affected by the initial scores of postnatal attachment, and whether attachment development is moderated by initial reports of postnatal child temperament and mothers’ depressive symptoms. Methods: All current data are based on data from the Norwegian Mamma Mia-study, where 609 perinatal women were included. A RCT-study was conducted, comparing the effects of Mamma Mia on attachment between an intervention and a control group at four measuring points after giving birth. Attachment was measured by a short version of the Parental Stress Index. The data was analyzed using independent samples t-test and mixed design ANOVA. Results: Results showed a consistent tendency of healthier attachment-scores in the intervention group, but the differences between groups were not statistically significant. Initial reports of both postnatal attachment, child temperament and mothers’ depressive symptoms were significantly related to attachment development over time, but with no difference between the intervention and control groups. Conclusion: The consistency of healthier scores on attachment in the Mamma Mia- group, suggests an effect of the intervention on attachment which may be of clinical relevance, if not statistically significant.